A novel text clustering algorithm treated attributes differently

But the traditional partitional clustering algorithms represented by C-means and fuzzy C-means take the same assumption that the attributes of objects play the same role in clustering. This is not desirable in documents clustering. Sometimes, the part attributes contribute more than others in deciding the cluster structure. But how to distinguish the importance of these attributes? This paper proposes a new fuzzy c-means text clustering algorithm. This algorithm can remark the importance of each attribute at the same time to realize the soft partition. It can find the real cluster structure hiding by the noises data.